Mitochondrial dysfunction has long been implicated in the pathogenesis of PD. However, the exact mechanisms by which mitochondrial dysfunction arises in the disease process and how cells, particularly neurons, handle dysfunctional mitochondrial are not well understood. The identification of a mitochondrial quality control system involving two FPD genes, PINK1 and Parkin, has provided a much-needed point of entry to elucidate the role of mitochondria in the pathogenesis of PD. Here we showed that PINK1 directly regulates mitochondrial transport and that it affects the stability and/or activity of Miro, a mitochondrial Rho GTPase with a well-establish function in mitochondrial transport. Our conclusion is supported by the following evidence: 1) dMiro protein level is negatively regulated by PINK1 and Parkin in vivo in Drosophila; 2) Overexpressed PINK1 and Parkin act together to promote the ubiquitination and degradation of hMiro1 in HeLa cells; 3) Reduction of the activities of Miro or other components of the mitochondrial transport machinery effectively rescued dPINK1 mutant phenotypes. 4) Overexpression of dMiro in DA neurons phenocopied dPINK1 loss-of-function effects; 5) Manipulation of dPINK1 activity produced clear mitochondrial motility phenotypes opposite to that observed for dMiro manipulation in Drosophila larval motor neurons. Together, these results support that the mitochondrial transport defects caused by PINK1 inactivation represent one of the key pathogenic events that contribute to PD pathogenesis in the Drosophila model.
Neurons are highly polarized cells that rely heavily on axonal transport to distribute to axons and synapses critical proteins and organelles synthesized in the cell body, thereby maintaining neuronal function and health. Defects in axonal transport are often linked to diseases affecting peripheral neurons that tend to extend very long axons 
. Although the symptoms of PD patients mainly arise from the loss of DA neurons, some PD patients carrying particular PINK1
mutations developed peripheral neuropathy with unknown cause 
. Our results showing the PINK1/Parkin pathway playing a critical role in regulating mitochondrial transport offers one potential explanation of the peripheral neuropathy symptoms observed in these PINK1 or Parkin-linked PD cases. It would be interesting to examine whether Miro protein level or activity is affected by these particular mutations in human cells. Moreover, we propose that defects in PINK1/Parkin-regulated mitochondrial transport may offer one explanation of the selective vulnerability of DA neurons observed in PD patients and animal models. DA neurons that make elaborate and long projections may be particularly vulnerable to impairment of the mitochondrial transport system.
Our results offer new insights into the mode of action of the PINK1/Parkin pathway in mitochondria quality control. We showed that, in Drosophila
models, PINK1 OE led to decreased mitochondrial flux and net velocity, as observed in Miro knockdown background. In addition, we found that Miro knockdown could facilitate an early step of mitophagy in mammalian cells. These observations, together with the finding that the normally labile PINK1 protein is stabilized on damaged mitochondria 
, suggest a scenario whereby the accumulation of PINK1 on damaged mitochondria and the subsequent turnover of Miro could exert neuroprotection by (1) preventing damaged mitochondria from being anterogradely transported along the axons, thus increasing their chance of getting eliminated in the soma; and (2) promoting elimination of damaged mitochondria through mitophagy. This potentially explains the normal protective function of PINK1. When PINK1 function is impaired, however, on one hand mitochondria become dysfunctional as evidenced by morphology changes and impaired electron transport chain function 
, on the other hand, the anterograde mitochondrial transport is enhanced as shown in this study in Drosophila
models. As a result, the dysfunctional mitochondria would have increased retention in the axons and synapses, resulting in increased reactive oxygen species (ROS) production, oxidative damage, and subsequent synaptic and axonal degeneration and eventual neuronal loss, at least in the Drosophila
models. Many details of this model await further experimental validation. For example, it has been suggested that the reported effect of PINK1/Parkin on mitochondrial autophagy may not operate in the same manner in primary neurons as compared to cultured non-neuronal cells 
. It also remains to be determined whether the effects of Miro on mitochondrial transport and mitophagy reflect a functional antagonism between these two processes, or two distinct functions of Miro in neuronal maintenance. In this respect, it is worth noting that the effect of Miro overexpression on cell survival in Drosophila
is cell type-dependent: it causes DA neuron loss but has no obvious effect on muscle integrity (data not shown). It is possible that different tissues may have different sensitivities to impairments of Miro function. For example, muscle cells may be less susceptible to mitochondrial transport defects than neurons.
One interesting difference we observed between Drosophila
and mouse systems was that although activation of the PINK1/Parkin pathway led to reduced Miro protein level in both systems, the loss of PINK1 in Drosophila
resulted in increased steady-state Miro protein level, whereas its loss in mammalian cells as in PINK1 (−/−)
MEF cells or PINK1 RNAi HeLa cells had the opposite effect. The mechanism of Miro downregulation in PINK1 loss-of-function mammalian cells is currently unknown, but it is presumably different from that used by activation of the PINK1/Parkin pathway. Since the upregulation of dMiro in dPINK1
mutant background is likely causal to DA neuron degeneration, as indicated by the rescue of DA neuron loss in dPINK1
mutant by dMiro-RNAi and the induction of DA neuron loss by dMiro-OE alone, it is tempting to speculate that the downregulation of Miro levels in PINK1 (−/−)
mouse, as opposed to the dMiro upregulation in Drosophila PINK1
mutant, might contribute to the lack of DA neuron degeneration phenotype in the mouse PINK1 models 
. Testing this hypothesis will require in vivo
studies boosting Miro expression levels in wild type and PINK1 (−/−)
Our results also provide new insights into the process by which the PINK1/Parkin pathway promotes mitophagy. Previous studies suggested that upon recruitment to damaged mitochondria, Parkin activates the ubiquitin proteasome system to effect wide-spread degradation of OMM proteins in an autophagy-independent manner, and it was further proposed that this remodeling of OMM is important for a subsequent step of mitophagy 
. The previously identified Parkin substrates, Mfn1 and Mfn2, although important for the effect of Parkin on mitochondrial fission/fusion dynamics, are not necessary for Parkin-induced mitophagy 
. Here we show that removal of Miro by the PINK1/Parkin pathway, in a presumably autophagy-independent but ubiquitination-dependent manner, facilitated mitophagy. Interestingly, knockdown of mammalian Miro itself promotes the formation of ring-like or round-shaped mitochondrial morphology, which is often observed in depolarized, mitophagy-ready mitochondria (; 
). It is possible that the removal of Miro from OMM exposes certain recognition signals for the autophagy machinery, or that Miro/Milton/Kinesin-mediated mitochondrial transport may normally antagonize the mitophagy process. Supporting the latter scenario, an interaction between the BECLIN 1-interacting protein AMBRA1 and the dynein motor complex has been implicated in mammalian autophagy 
. It also remains to be understood at the mechanistic level how PINK1 cooperates with Parkin to promote the ubiquitination and degradation of Miro. One attractive hypothesis is that PINK1 may directly phosphorylate Miro to promote its subsequent ubiquitination and degradation by Parkin, as suggested by a recent study 
. However, our biochemical data have so far failed to support this hypothesis. It is possible that the divergent results are due to the different cell lines used or other experimental conditions. Alternatively, PINK1 may directly act on Parkin to promote Parkin's mitochondrial recruitment or activity in activating ubiquitin proteasome system-mediated ubiquitination and degradation of Miro. Further studies are needed to elucidate the molecular mechanisms of PINK1/Parkin action.
Finally, it is worth mentioning that studies in Drosophila
models have identified a number of genetic modifiers of PINK1/Parkin 
. While some of these genetic modifier genes may be directly related to the seemingly diverse biological activities of the PINK1/Parkin pathway, possibly mediated by distinct PINK1/Parkin substrates, others may reflect cellular compensatory responses to cope with the mitochondrial dysfunction caused by PINK1/Parkin inactivation. The fact that manipulations of each of these different cellular processes exert clear functional rescue of PINK1/Parkin mutant phenotypes suggests that there exists a signaling network linking the diverse activities of PINK1/Parkin in mitochondria biology with the nuclear-encoded cellular responses to mitochondrial dysfunction, and that many key players in this network represent novel and rational therapeutic targets.